A Quality-Improvement Project Use of a Computerized Medical Database and Reminder Letters to Increase Preventive Care Use in Kaiser Permanente Patients

Abstract

Context: Previous studies have suggested that preventive health care measures may be improved by proactive patient-reminder systems and use of electronic medical databases.

Objective: Our objective was to use Kaiser Permanente's (KP's) electronic medical databases to improve the preventive health care delivered to KP patients in Honolulu, HI.

Design: Patients not seen by their primary care physician in more than one year (“low-utilizing patients”) and overdue for preventive health care services were identified using KP's electronic medical databases. These patients were then sent letters requesting that they obtain these services. Response rates and results of their screening tests were tracked.

Results: Of 309 letters sent, 84 (27.2%) patients responded. Of the 260 patients who were reminded of overdue immunizations (tetanus, pneumonia, influenza, or a combination of these), 51 (19.6%) came in to obtain them. Ten of the 37 (27.0%) women overdue for breast cancer screening came in for mammograms. Eleven of the 109 (10.1%) patients overdue for colorectal cancer screening completed fecal occult blood testing.

Conclusion: Outreach letters that target low-utilizing patients identified by an electronic medical database may be an efficient and cost-effective way of improving patient use rates of preventive health care.

Context

Preventive medicine is the forefront of medical care. It is cost-effective, improves morbidity, and reduces mortality. However, it is also often overlooked. A report of a study of 12 metropolitan areas published in 2003 noted that on average, US patients receive only 55% of recommended preventive care.¹ Since then, various strategies have been pursued to improve outreach to patients to address preventive-care deficiencies.

One example is the use of patient-reminder systems, such as phone calls or letters. These have been successfully used to improve preventive care such as vaccinations,^2–4 cancer screening,^5–12 smoking cessation,¹³ and diabetes management.^14,15 However, although mass mailings or phone calls can target large volumes of patients, they are impersonal and cannot be tailored to address multiple deficits in preventive care without individual analysis of each patient's record, which would be time-consuming and impractical for a busy practice or large health maintenance organization to manage. Thus, generic patient-reminder systems can address preventive care only on a population level, not at the individual level.

Electronic medical records (EMR) have also been successfully implemented as another tool to improve preventive medicine.^16,17 By consolidating patient records into a single database, clinicians can easily access and reference information about preventive care. This allows physicians to easily address patient deficiencies in preventive care with minimal effort. Other aspects of EMR, such as computerized reminders about preventive-care deficiencies have also been used to improve compliance with treatment plans.^18–22 Providing automatic reminders to clinicians immediately— while they are seeing each patient—allows them to address individual deficits. However, compliance with these reminders has been limited because of various barriers, including time constraints in busy clinics.^23–27 Additionally, deficits in preventive care are not usually addressed unless a patient sees his or her physician. Thus, EMR and computerized reminder systems are typically limited to addressing preventive care on an individual level and cannot efficiently address populations of patients with preventive-care deficiencies.

… the PMST allows physicians to address the health of their patients at both a population level and an individual level.²⁸

In an effort to address preventive medicine at a population level while addressing each patient as an individual, a group of Kaiser Permanente (KP) physicians designed the Panel Management Support Tool (PMST). The PMST is a Web-based computerized member database designed to complement the existing KP EMR database (KP HealthConnect). By offering summary information for each primary care physician's (PCP's) panel of patients, it allows for specific targeting of defined subsets of patients in each panel. It can also identify specific deficits in preventive care for each individual patient. Used in a proactive manner, the PMST allows physicians to address the health of their patients at both a population level and an individual level.²⁸

Objective

We sought to use the PMST to improve the level of preventive health care delivered to KP patients in Honolulu, HI.

Methods

Design Overview

Four internal medicine residents, under the supervision of a staff physician at the KP Honolulu clinic, piloted a quality-improvement project designed to use the PMST. Each resident analyzed their preceptor's panel with the PMST. Patients who had not been seen by their PCP in more than one year (“low-utilizing patients”) were first identified. The patients were then screened by the PMST for overdue care on the basis of national recommendations for preventive health care services (eg, vaccinations, screening laboratory tests, cancer screening). Targeted patients were sent a personalized reminder letter requesting that they come in to receive overdue health care services. Patient responses were subsequently tracked to determine the efficacy of this strategy.

Setting and Participants

The project was conducted from September 2006 through March 2007. Panels for four KP internal medicine physicians, all based at the Honolulu clinic (an urban clinic), were used. Panel sizes ranged from 783 to 1799 patients, with a total of 5196 patients initially screened. (Range in panel sizes was due to clinic adjustments for practitioner employment level, from full time to half time.) Using the PMST, patient panels were divided by month of birth (September through December birthday months were randomly targeted), for ease of approach for initial screening, for a total of 1440 patients. Patients seen by a PCP within one year were then excluded. The PMST then identified 309 remaining patients who did not meet screening recommendations for hyperlipidemia, diabetes, kidney function, mammography, or fecal occult blood testing according to national guidelines, as well as recommended immunizations for tetanus, pneumonia, and/or influenza vaccination. These showed up as a “care gap” in the PMST.

No review by an institutional review board was done, because this study was an internal quality-improvement project.

Design

As mentioned, we first used the PMST to identify low-utilizing patients. Preventive services targeted included screening for cholesterol level and glucose level; kidney function testing; tetanus, pneumonia, and/or influenza vaccines; and mammograms and/or fecal occult blood testing.

We then did a brief chart review to exclude patients who had an upcoming medical appointment or who had been recently contacted to come in for health care. Letters were sent to the identified patients, notifying them of overdue services and requesting that they come in to obtain these services within the next month. Chart review and letter preparation took approximately four to five minutes per patient. A response period of three months was allowed, randomly chosen to show a temporal relationship between the outreach letters being sent out and the patients coming in for care.

Laboratory test results were communicated by letter or phone call, by the residents, under the supervision of a KP staff physician, to the patients who came in response to the outreach letters. Abnormal laboratory test values were defined as any test value outside of the normal value range. Educational booklets (about high cholesterol, prediabetes, etc) were sent to patients with abnormal test values. Medications were initiated and follow-up testing and referrals were recommended if deemed appropriate.

Patient responses were tracked as results were received. Patient encounters were documented in the KP medical database so that the primary care teams would be aware of the outreach efforts and could follow up regarding any test results.

Results

Patients' responses to the letters were tracked and examined in relation to age and sex (Table 1), type of care requested, and clinician panel size. Of those targeted patients, 160 of 309 (52%) were older than 50 years and 149 (48%) were younger, with the mean age being 49.2 years. The median age was 50 years (range, 22–90 years). Of the 309, 175 (57%) were women and 134 (43%) were men.

Table 1.

Response rates by age and sex

	Age ≥ 50 years	Age < 50 years	Women	Men
Total number of letters sent	160	149	175	134
Total number of responses	48	35	56	27
Percentage of responses of total sent	30	23	32	20

Of the identified patients, 128 were overdue for screening blood work, 260 for vaccines, and 122 for cancer screening (mammogram and/or fecal occult blood testing). Most patients were overdue for a combination of these. Of the 309 patients to whom letters were sent, 84 (27.2%) responded (Table 2): 42 came in for blood work, 51 came in to receive recommended vaccinations, and 21 completed requested breast cancer screening or initial colon cancer screening (not equal to 84 because some patients received more than one type of care). Of the 260 patients who were reminded of overdue immunizations (tetanus, pneumococcal, influenza, or a combination of these), 51 (19.6%) came in to have them completed. Ten of the 37 women (27.0%) overdue for breast cancer screening came in for mammograms, and 11 of the 109 patients (10.1%) overdue for colorectal cancer screening completed fecal occult blood testing.

Table 2.

Response rates by tests or treatments

Overdue care	Total number of tests/vaccines ordered	Total responded(%)
Screening laboratory tests
Lipid profile	120	40 (33.3)
Fasting glucose	36	27 (75)
Renal function	20	8 (40)
Immunizations
Tetanus vaccine	240	45 (18.8)
Pneumococcal vaccine	31	4 (12.9)
Influenza vaccine	36	6 (16.7)
Cancer screening
Mammography	37	10 (27.0)
Fecal occult blood testing	109	11 (10.1)

Four different PCP panels, ranging from 783 to 1799 patients, were screened by the PMST. Response rates from each clinician ranged from 20.7% to 38.2% (Table 3).

Table 3.

Response rates by physician

Physician	Panel size	Number of letters senta	Number of responses (%)
1	783	82	17 (20.7)
2	1799	89	34 (38.2)
3	1456	91	20 (22)
4	1169	47	13 (27.7)
Total	5196	309	84 (27.2)

^a Letters sent between October and December 2006.

Discussion

This quality-improvement project used the PMST to specifically target low-utilizing patients. By doing this, we focused preventive care on patients who might never have otherwise sought it. Using a reminder letter, we obtained an overall response rate of 27.2%, which is comparable to that of other studies using such letters (range, 19%–69%).^{2–4,7,9,11,12,15,29,30}

Aside from providing the benefits of routine screening, outreach letters might also improve future medical care for low-utilizing patients. A meta-analysis by Cabana and Jee³¹ of five studies examining preventive care demonstrated improved outcomes for screening with sustained continuity of care. Thus, inducing low-utilizing patients to see their PCP again could potentially increase future compliance with preventive-care plans.

The potential ease of use by medical support staff would also make preventive care efforts cost-effective and less dependant on the individual PCP.

The PMST also identified patients who were arbitrarily assigned to a PCP's panel but had never seen their PCP before. A study by Zhu et al³² showed that patients sent reminder letters for health screening who saw physicians new to them had odds ratios six- to eightfold higher for obtaining health screenings than patients who did not see a physician. Therefore, targeting this population of patients could result in improved screening for preventive care.

The PMST screened multiple areas of preventive care, including screening laboratory tests, immunizations, and cancer screening. This allowed multiple deficiencies to be addressed in a single letter. A study by Burack et al³³ showed that the combination of reminders for Papanicolaou smears and mammograms had an independently better effect on cervical cancer screening than did a reminder for a Papanicolaou smear or a reminder for a mammogram alone.³³ Another study by Terrel-Perica et al³⁰ showed that patients who received a combination letter for both influenza and pneumococcal immunizations were more likely to receive simultaneous immunization than were patients who received a letter for only influenza vaccination or no letter at all. Thus, targeting multiple deficiencies in different areas of preventive medicine might have improved patients' motivation to address their preventive-care needs.

The PMST was used to screen the differently sized panels of four clinicians of various experience levels, with response rates of 20% to 38%. Although various studies have shown that clinician characteristics can be a barrier to certain types of preventive care,^34–36 these response rates demonstrate that different providers can use the PMST effectively for preventive care.

Finally, although resident physicians performed this project, the PMST is relatively simple to use. By delegating screening to medical support staff members, PCPs can address more preventive care needs. In a study by Denberg et al,³⁷ 93% of patients reported that they believed that PCP involvement in preventive health care is not always necessary, is inconvenient, or represents an unnecessary expense, and more than 70% said that they were open to a non-PCP-centered method of receiving preventive services. Our design was simple and easily reproducible and could easily be implemented by medical support staff.

Our results show that an EMR database program such as the PMST is a useful tool for increasing patient use of preventive health care. By specifically designing outreach letters for individual deficiencies in preventive care, clinicians can attempt to ensure adequate preventive-care measures for all of their patients, rather than only the patients who frequently come in to receive health care. The potential ease of use by medical support staff would also make preventive-care efforts cost effective and less dependant on the individual PCP.

Thus, the PMST would allow PCPs to reach out to large populations of patients deficient in multiple aspects of preventive care with minimal cost and effort.

Future Directions

The protocol we used was simple enough to be carried out by medical support staff and would be easy to transition to a larger scale. This project could be expanded to include the entire Honolulu clinic or even the entire region. A larger, prospective, randomized study might be considered to further evaluate efficacy and outcomes.

In addition, a more detailed assessment of laboratory value or screening abnormalities might give further information about the effect of this type of proactive preventive care on patients. For example, a diagnosis of overt diabetes, instead of prediabetes, would likely have a much more significant effect on the projected morbidity and mortality of a low-utilizing patient. Similarly, benefits of renal function monitoring might have more of an effect in patients with a lower glomerular filtration rate than in those with healthier kidneys.

To further evaluate the effect of the outreach letters, simple follow-up surveys could be sent to the targeted patients to assess whether the letters influenced their health awareness, views of KP, or perception of their overall health.

Conclusion

Using an electronic medical database that can target low-utilizing patients and reminder letters may improve preventive health care offered to KP patients.

Disclosure Statement

The author(s) have no conflicts of interest to disclose.

Acknowledgments

Katharine O'Moore-Klopf, ELS, of KOK Edit provided editorial assistance.

Modern Life

The overwhelming importance of preventive medicine, sanitation, and public health indicates that in modern life the medical profession is an organ differentiated by society for its own highest purposes, not a business to be exploited by individuals according to their own fancy.

— Medical education in the United States and Canada, Abraham Flexner, 1866–1959, American Educator